Site-Directed Mutagenesis of Firefly Luciferase Active Site Amino Acids: A Proposed Model for Bioluminescence Color

Page created by Marc Mclaughlin
 
CONTINUE READING
Biochemistry 1999, 38, 13223-13230                                              13223

     Site-Directed Mutagenesis of Firefly Luciferase Active Site Amino Acids: A
                    Proposed Model for Bioluminescence Color†
 Bruce R. Branchini,* Rachelle A. Magyar, Martha H. Murtiashaw, Shannon M. Anderson, Lisa C. Helgerson, and
                                                Marc Zimmer
                         Department of Chemistry, Connecticut College, New London, Connecticut 06320
                               ReceiVed May 21, 1999; ReVised Manuscript ReceiVed July 22, 1999

        ABSTRACT: Under physiological conditions firefly luciferase catalyzes the highly efficient emission of
        yellow-green light from the substrates luciferin, Mg-ATP, and oxygen. In nature, bioluminescence emission
        by beetle luciferases is observed in colors ranging from green (∼530 nm) to red (∼635 nm), yet all
        known luciferases use the same luciferin substrate. In an earlier report [Branchini, B. R., Magyar, R. M.,
        Murtiashaw, M. H., Anderson, S. M., and Zimmer, M. (1998) Biochemistry 37, 15311-15319], we
        described the effects of mutations at His245 on luciferase activity. In the context of molecular modeling
        results, we proposed that His245 is located at the luciferase active site. We noted too that the H245
        mutants displayed red-shifted bioluminescent emission spectra. We report here the construction and
        purification of additional His245 mutants, as well as mutants at residues Lys529 and Thr343, all of which
        are stringently conserved in the beetle luciferase sequences. Analysis of specific activity and steady-state
        kinetic constants suggested that these residues are involved in luciferase catalysis and the productive
        binding of substrates. Bioluminescence emission spectroscopy studies indicated that point mutations at
        His245 and Thr343 produced luciferases that emitted light over the color range from green to red. The
        results of mutational and biochemical studies with luciferase reported here have enabled us to propose
        speculative mechanisms for color determination in firefly bioluminescence. An essential role for Thr343,
        the participation of His245 and Arg218, and the involvement of bound AMP are indicated.

   Bioluminescence is an appealing process by which living          According to the original theory (5, 6) based predominantly
organisms convert chemical energy into light. Beetle biolu-         on studies of the P. pyralis enzyme, red light emission (λmax
minescence, observed in perhaps 2000 species of fireflies           615 nm), which is observed at pH ∼6.0, results from the
(1, 2), has provided a fruitful area of basic and applied           keto form of the emitter. At pH 7.8, the familiar yellow-
research mainly focused on the North American firefly               green light emission (λmax 560 nm) is produced from the
Photinus pyralis (3). In this firefly species, the luciferase-      enolate dianion form of the excited state oxyluciferin by a
catalyzed light emission process is extremely efficient; in         presumed enzymatic assisted tautomerization (step f). In
alkaline solution, nearly a single photon is emitted per reacted    nature, beetle luciferases, which all utilize the same firefly
luciferin molecule (4). The enzyme firefly luciferase func-         luciferin substrate, display various colors of light from green
tions as a monooxygenase, without the apparent involvement          (λmax ∼530 nm) to red (λmax ∼635 nm) (7, 8).
of a metal or cofactor.                                                Color modulation has also been attributed to alterations
   According to the enzymatic mechanism (5, 6) presented            in firefly luciferase structures resulting in changes of polarity
in Figure 1, the beetle luciferases catalyze a sequence of          (9) or rigidity (10, 11) of the emitter binding site. In the
reactions that convert firefly luciferin into an electronically     latter theory (10, 11), conformations of the keto form of
excited state oxyluciferin product which then emits light.          oxyluciferin, related by rotation about the C2-C2′ bond
First, firefly luciferase catalyzes the formation of an enzyme-     (Figure 1), account for the color variation. Red emission is
bound luciferyl adenylate (Figure 1, step a). Next, a proton        attributed to the minimum energy conformation of an unusual
is abstracted from the C-4 carbon of the adenylate by a             twisted intramolecular charge transfer (TICT)1 excited state
presumed enzyme base (step b); molecular oxygen adds to             (Figure 1) in which the aromatic rings of oxyluciferin are
the newly formed anion (step c); and an electronically excited      rotated 90° about the C2-C2′ bond. Yellow-green light
state oxyluciferin molecule and CO2 are produced from a             emission would result from a higher energy excited state
highly reactive dioxetanone intermediate (steps d and e).           conformer. The proposed mechanisms for bioluminescence
                                                                    color have been reviewed (10-12).
   †
     This work was supported by a grant from the National Science
Foundation (MCB 9816898) and the Hans & Ella McCollum-Vahlteich
’21 Endowment. Portions of this work were presented at the 10th        1 Abbreviations: BPTC, 2-(4-benzoylphenyl)thiazole-4-carboxylic

International Symposium on Bioluminescence and Chemiluminescence,   acid; CB, 50 mM Tris-HCl (pH 7.0), 150 mM NaCl, 1 mM EDTA,
Bologna, Italy, September, 1998.                                    and 1 mM DTT; GST, glutathione-S-transferase; PheA, phenylalanine-
   * To whom correspondence should be addressed at the Department   activating subunit of gramicidin synthetase 1; SDS-PAGE, sodium
of Chemistry, Connecticut College, 270 Mohegan Ave., New London,    dodecyl sulfate-polyacrylamide gel electrophoresis; TICT, twisted
CT 06320. Telephone: (860) 439-2479. Fax: (860) 439-2477. E-        intramolecular charge transfer; WT, wild-type Photinus pyralis lu-
mail: brbra@conncoll.edu.                                           ciferase containing the additional N-terminal peptide GPLGS.

                              10.1021/bi991181o CCC: $18.00 © 1999 American Chemical Society
                                                 Published on Web 09/14/1999
13224 Biochemistry, Vol. 38, No. 40, 1999                                                                           Branchini et al.

FIGURE 1: Mechanism of firefly luciferase catalyzed bioluminescence and TICT excited state of oxyluciferin.

   Mutational studies of firefly luciferases have been con-             Previously, we had reported (27) that 2-(4-benzoylphenyl)-
ducted (12-18) that have identified changes in single and/           thiazole-4-carboxylic acid (BPTC) was a potent active site-
or groups of amino acids widely distributed throughout the           directed photoinactivation reagent for P. pyralis luciferase.
sequences of nine luciferases that affect the color of               The degradation of a luciferase peptide, 244HHGF247, was
bioluminescence. The comparative studies of these luciferase         directly correlated to the photoinactivation process. The
sequences from five firefly speciessHotaria parVula (15),            results of a subsequent mutagenesis study (25) established
Pyrocoelia miyako (15), Luciola cruciata (14), Luciola               His245 as the likely primary target of BPTC-catalyzed
mingrelica (16, 17), and P. pyralis (13)sand the four click          enzyme inactivation. According to our proposed molecular
beetle [Pyrophorus plagiophthalamus (12, 18)] isozymes have          model (25), helix residues 245HGF247 comprise part of the
not yielded any direct structure-function relationship to            luciferase binding pocket for luciferin. The His245 side-chain
account for the color variation of firefly bioluminescence.          imidazole group is likely positioned at the opening of this
   The cloning and sequencing of P. pyralis luciferase and           binding pocket. However, the His245 side chain was not
similar enzymes from several beetle species (12, 19) have            found to be an essential group for productive luciferin
revealed that the luciferases are closely related to a large         binding or efficient light production. Interestingly, all of the
family of nonbioluminescent proteins (20, 21) that catalyze          His245 mutants catalyzed the emission of light differing from
reactions of ATP with carboxylate substrates to form acyl            the characteristic yellow-green of native and wild-type (WT)
adenylates. The P. pyralis luciferase crystal structure without      P. pyralis luciferase. Since it is very likely that oxyluciferin
bound substrates (22), the first one from the “acyl-adenylate/       binds similarly to luciferin, we used our molecular model to
thioester-forming” enzyme family (23), revealed a unique             determine that, in addition to His245, the side-chain groups
molecular architecture consisting of a large N-terminal              of residues Thr343 and Lys529 also might be in close
domain (residues 1-436) and a small C-terminal domain                proximity to the thiazolinone ring of the emitter. We report
(residues 440-550). A general location of the P. pyralis             here the results of new mutational studies of His245, Thr343,
                                                                     and Lys529 directed at the mechanistic question of color
luciferase active site was proposed (22) based on an analysis
                                                                     determination in firefly bioluminescence.
of the positions of conserved residues among approximately
40 enzymes sharing the adenylation function. However, the            MATERIALS AND METHODS
exact locations of the substrate binding sites could not be             Materials. P. pyralis cDNA in the vector pSx(T7)-Ppy was
established (22). Subsequently, the crystal structure of a           a generous gift of Monika Gruber. The following items were
second member of the adenylate family, the phenylalanine-            obtained from the indicated sources: Mg-ATP (equine
activating subunit of gramicidin synthetase 1 (PheA) in a            muscle) (Sigma); luciferin (Biosynth AG); restriction endo-
complex with Phe, Mg ion, and AMP, was reported (24).                nucleases, T4 polynucleotide kinase, and T4 DNA ligase
Starting with the crystal structures of P. pyralis luciferase        (New England Biolabs); and mutagenic oligonucleotides
(22) and PheA (24), we used molecular modeling techniques            (Genosys). Glutathione-S-transferase (GST) fusion constructs,
to produce a potential model of the active site containing           encoding WT and mutant fusion proteins, were expressed
substrates luciferin and Mg-ATP (25). Recently, the crystal          and purified as previously described (25).
structure of P. pyralis luciferase containing two molecules             General Methods. Firefly luciferase bioluminescence
of bromoform, a general anesthetic and luciferease inhibitor,        activity was determined using flash height-based light assays.
was described (26). One of the bromoform molecules binds             The standard activity assays were performed in 25 mM
in a site tentatively identified as the luciferin binding pocket,    glycylglycine buffer (pH 7.8) as described previously (25,
based on the location of the phenylalanine binding site in           28-30). All flash height measurements were made by
the PheA structure (24, 26). The newly proposed (26)                 injecting the initiating solutions with a Hamilton Microlab
luciferin site binding is in good general agreement with our         900 automatic dispenser operated at the maximum speed.
proposal (25).                                                       The estimated mixing time was ∼0.2 s. Data were acquired
Firefly Bioluminescence Color                                                   Biochemistry, Vol. 38, No. 40, 1999 13225
with an SLM-Aminco Chem-Glow II Photometer interfaced            “phosphorescence” mode, and dark current is subtracted prior
to a Strawberry Tree, Inc., A/D converter operated at a          to the initiation of the bioluminescence reactions.
sampling rate of 0.05 s. All light measurements were                Reaction mixtures containing WT or mutant luciferases
corrected for the spectral response of the Hamamatsu 931B        (2-300 µg) in 0.2 mL of CB (0.3 mL of CB for the T343
photomultiplier tube used in the assays. The corrections         mutants and a WT control), luciferin, and Mg-ATP (Table
applied were based on the corrected bioluminescence emis-        1) were brought to a final volume of 0.5 mL with 25 mM
sion spectra obtained as described below. Emission spectra       buffer: MES (pH 5.5-6.5); glycylglycine (pH 7.8); or Tris
for previously reported P. pyralis luciferase mutants (25)       (pH 8.0-9.0). The pH of the mixtures was verified after each
were remeasured, and specific activity data were recalculated,   emission spectrum was measured. Approximately 2 min after
resulting in several minor changes (Table 1).                    initiation of bioluminescence, spectra were recorded in a 1.0
   The following measurements were performed as described        mL optical glass fluorescence cell at a scan rate of 60 nm/
previously: guanidinium chloride-induced denaturation of         min. Separate measurements of emission intensity for each
the luciferases monitored by fluorescence spectroscopy (25,      reaction indicated decreases of no more than ∼5% during
31); circular dichroism spectroscopy (25); and steady-state      the recording interval. All emission spectra were corrected
kinetic constants for luciferin and Mg-ATP (25, 29, 30).         for the spectral response of the Perkin-Elmer R928 photo-
   Site-Directed Mutagenesis. Mutations at H245, K529, and       multiplier tube employed, and these spectra were used to
T343 were created by oligonucleotide-directed mutagenesis        determine the spectral corrections for data collected with the
(32) as previously described (25). The following mutagenic       light assay device described above.
primers were designed for annealing onto the ssDNA                  Molecular Modeling. The starting conformation of oxy-
template: H245R, 5′-T TTA AGT GTG GTA CCA TTC                    luciferin and AMP within P. pyralis luciferase was deter-
CAT CGC GGT TTT GG-3′[KpnI]; H245Q, 5′-T TTA AGT                 mined by modifying the previously reported lowest energy
GTG GTA CCA TTC CAT CAG GGT TTT GG-3′ [KpnI];                    conformation of luciferase in a complex with luciferin, ATP,
H245N, 5′-T TTA AGT GTG GTA CCA TTC CAT AAC                      and Mg2+ (25). The low-energy conformations used to draw
GGT TTT GG-3′[KpnI]; K529R, 5′-GGT CTT ACC GGT                   Figure 6 were generated by conducting Monte Carlo searches
AGG CTC GAC GCA A-3′ [AgeI]; K529Q, 5′-GTA CCG                   on the starting complex. Two searches of 5000 Monte Carlo
AAA GGG TTA ACC GGA CAA CTC GAC GCA-3′                           steps each were conducted as described earlier (25) using
[HpaI]; T343S, 5′-A GGG ATA CGC CAA GGA TAT GGG                  the following: a “hot” sphere of 12 Å around the oxylu-
CTC TCT GAG ACT ACA TCA-3′ [StyI]; T343A, 5′-A                   ciferin; short restraints between the oxyluciferin phenolate
GGG ATA CGC CAA GGA TAT GGG CTC GCT GAG                          ion and two of the side-chain Arg218 NH’s (desired distance
ACT ACA-3′ [StyI] (underline represents silent changes to        2.5 Å with a half-width flat bottom of 1.5 Å and a force
create a unique screening endonuclease site, boldface rep-       constant of 200 kJ/Å2); and translation and rotation of the
resents the mutated codon, and brackets indicate the endo-       oxyluciferin by between 0 and 1.00 Å and 0 and 180° per
nuclease used). The primer 5′-CAC ACA GTA GAC AGG                step. Conformational families were determined by cluster
ATC ATG GAA GAC G-3′ (underline represents a new AccI            analysis and by filtering hydrogen bonding interactions in
site) was designed to align the WT cDNA reading frame for        the active site.
subcloning (25) into the pGex-6p-2 fusion vector and was
included in each mutagenesis reaction. The entire cDNA of        RESULTS
all luciferase genes was verified by DNA sequencing.                OVerexpression, Purification, and Characterization of
   Protein Analysis. Luciferase protein concentrations were      Luciferase Proteins. WT and mutant luciferases listed in
determined by UV spectroscopy using 278 ) 45 560 M-1            Table 1 were expressed as GST-fusion proteins, which, after
cm-1 (33). For solutions containing DTT, the Bio-Rad             protease cleavage, contained the additional N-terminal pep-
Protein Assay was used with BSA as the standard. SDS-            tide GPLGS. Proteins were purified to homogeneity (as
PAGE was performed according to the method of Laemmli            judged by SDS-PAGE) in yields of 8-14 mg/L of culture
(34), and gels were stained with Coomassie R-250.                volume. All enzymes maintained full activity for at least 10
   Bioluminescence Emission Spectra. Emission spectra of         months when stored at 4 °C in CB containing 0.8 M
bioluminescence reactions were obtained over the pH range        ammonium sulfate and 2% glycerol.
5.5-9.0 using a Perkin-Elmer LS-5B luminescence spec-               The far-UV circular dichroic (CD) spectra (data not
trometer operated in the “phosphorescence” mode with the         shown) of the His mutants and WT were nearly indistin-
excitation source shunted, the time delay set to 0.1 ms, and     guishable, indicating that mutations at position 245 had little,
the gate time set to 5.5 ms. For bioluminescence measure-        if any, effect on secondary structure. The tertiary structure
ments, the “phosphorescence” mode is more sensitive than         and thermodynamic stability of the H245 mutants appear to
the previously used (30) “fluorescence” mode. A pulsed           be very similar to those of WT, as judged by nearly
excitation source is activated, and gated detection periods      superimposable fluorescence emission spectra and identical
are used to accumulate signal intensity in both modes. In        guanidinium chloride-induced denaturation curves with the
the “fluorescence” mode, however, signals also are sampled       midpoint denaturant concentration equal to 3.5 ( 0.3 M (data
during a second gated detection period, which is coordinated     not shown). Similarly, the thermostabilities of the K529 and
to a time period when the excitation source is gated off. The    T343 mutants at 37 °C were within 10% of the value for
second measurement is subtracted from the first to remove        WT (data not shown).
background dark current. When measuring continuously                Bioluminescence ActiVity of the Luciferases. In an earlier
emitting bioluminescence signals in the “fluorescence” mode,     study (25) of the role of His245 in firefly bioluminescence,
the subtraction feature, which cannot be shut off, actually      we had prepared and characterized luciferase mutant proteins
subtracts real signal intensity, thereby lowering sensitivity    H244F, H245F, H245A, and H245D. Unlike the H244F
significantly. The subtraction feature is inactive in the        mutant, the His245 mutants displayed altered emission
13226 Biochemistry, Vol. 38, No. 40, 1999                                                                                Branchini et al.

Table 1: Specific Activity and Bioluminescence Emission of
Luciferase Enzymes
                                                   light emission
                       bioluminescence max (nm)b intensity ratioc
enzyme rel sp act.a       pH 7.8        pH 5.5       (green/red)
WT           100.0           558             613        3.1 (0.1)e
WTd          100.0           556             608        3.2 (0.5)e
H244F         18.4           557             612        3.7
H245R          0.05          579             618        0.9
H245F          8.0           595             620        0.6
H245A         26.4           604             620        0.5
H245Q          1.6           606             620        0.4
H245N         24.8           613             617        0.3
H245D          0.23          617             618        0.15
K529R          0.16          560             614        2.8
K529Q          0.06          560             610        2.5
T343Sd        26.4           560             617        2.7
T343Ad         1.1           617             621        0.1
   a Specific activity measurements were made as described previously

(25). WT values are defined as 100 and are equivalent to 1.18 × 1015
photons s-1/mg in flash height-based assays. Data were corrected for
the spectral response of the Hamamatsu 931B photomultiplier tube used
in the assays. In the standard activity assays, the final concentrations
of Mg-ATP and luciferin were 2 mM and 70 µM, respectively. For
mutants with elevated Km values, substrate concentrations were
adjusted: H245F and T343A, 3 mM Mg-ATP and 200 µM luciferin;
H245D, 4 mM Mg-ATP and 150 µM luciferin; and K529Q, 3 mM
Mg-ATP and 300 µM luciferin. b Bioluminescence emission spectra
were measured as described under Materials and Methods. c Ratio of
bioluminescence light emission intensity, 550/620 nm, at pH 7.8.
d Bioluminescence spectra were measured in higher ionic strength

solutions as described under Materials and Methods. e Measured at pH
5.5.

spectra compared to WT (Table 1, Figure 2). Our proposed
model (25) of the P. pyralis luciferase active site suggests
that only the side-chain groups of His245, Thr343, and                     FIGURE 2: Bioluminescence emission spectra of WT and the His245
Lys529 are within 6 Å of C4 or C5 of luciferin (Figure 1).                 mutants at pH 7.8. Conditions for recording and correcting the
                                                                           emission spectra are described under Materials and Methods.
Moreover, all 3 of these residues are conserved in all of the
17 published luciferase sequences (12, 18, 19). To evaluate                kinetic behavior of P. pyralis luciferase, the steady-state
the potential functional role of the three invariant residues              kinetic parameters for the WT and mutant enzymes were
in bioluminescence color, we expressed, purified, and                      determined. The H245D and H245F mutants exhibited
characterized the following: T343A and T343S; K529R and                    increases in the Km for luciferin of ∼4- and ∼6-fold,
K529Q; and three additional His245 mutant proteins (H245R,                 respectively, whereas the values for all of the other His245
H245Q, and H245N) (Table 1).                                               mutants were approximately the same as the WT value of
   Specific activities (photons s-1/mg) for all luciferases were           15 µM. The Km value of H245R for Mg-ATP was reduced
measured and expressed as flash height-based values (Table                 to about half that of the WT; the other H245 mutants had
1) that relate the maximum achievable overall reaction rate                elevated Km values ranging from an ∼1.5-fold increase for
for all steps shown in Figure 1. These specific activity values            H245A to an ∼11.5-fold increase for H245D (Table 2). For
are useful for making comparisons of the WT and mutant                     the position 529 mutants, the conservative substitution
properties. The specific activities of the H245A, H245N,                   K529R had almost no effect on the luciferin Km value and
H244F, H245F, and H245Q mutants were reduced ∼4-, ∼4-,                     resulted in an ∼4-fold reduction of the Mg-ATP Km value.
∼5.5-, ∼12.5-, and ∼62.5-fold, respectively, compared to                   Eliminating the positively charged side chain had substantial
WT (Table 1). The introduction of charged residues at                      effects on the Km values for both substrates, since the K529Q
position 245 had a greater effect on specific activity as the              mutant had ∼11-fold and ∼5-fold increased values for
values for H245R and H245D were reduced ∼2000- and                         luciferin and Mg-ATP, respectively (Table 2). The Km values
∼435-fold, respectively. For the position 529 mutants, the                 of the Thr343 mutants for luciferin were increased ∼1.5-
conservative substitution K529R caused a considerable                      and ∼6.5-fold with the Ser and Ala substitutions, respec-
∼625-fold lower specific activity compared to WT; however,                 tively. Both Thr343 mutants had increased Km values for
the loss of the positively charged side chain was far more                 Mg-ATP. T343S showed an ∼2.5-fold increase, while the
deleterious as the K529Q mutant had an ∼1650-fold reduced                  value for T343A increased ∼5.5-fold. For all of the luciferase
specific activity (Table 1). The conservative substitution of              enzymes except H244F, the effects of the single amino acid
Ser for Thr resulted in an ∼4-fold decrease, whereas the                   changes on kcat were greater than those on Km (Table 2).
removal of the hydroxyl functionality reduced the specific                 The kcat value of the H244F mutant decreased ∼7-fold while
activity of T343A by ∼90-fold.                                             the Km values for Mg-ATP and luciferin decreased ∼8.5-
   Effects of His245, Thr343, and Lys529 Substitutions on                  fold and 3-fold, respectively. All of the mutants are catalyti-
Substrate Binding and Catalytic Constants. To evaluate the                 cally compromised, and their kcat values are proportional to
effect of substitutions at positions 245, 343, and 529 on the              their respective specific activities (Table 2).
Firefly Bioluminescence Color                                                                  Biochemistry, Vol. 38, No. 40, 1999 13227

Table 2: Steady-State Kinetic Constants for Luciferase Enzymes                  H244F mutant contains the only substitution found among
                                                                                the luciferases (12, 19) that occurs in the four Pyrophorus
                             Km (µM)a
                                                                                isozymes (18). The H244F mutant had a modestly lower (∼5-
  enzyme         luciferin         Mg-ATP                  kcat (s-1)b          fold) specific activity; however, the Km values for both
  WT              15 ( 2           160 ( 20          1.6 ( 0.3                  substrates were reduced (Table 2). Compared to WT, the
  H244F            5(1              19 ( 2           0.22 ( 0.05                green light emitting click beetle isozyme also has reduced
  H245R           15 ( 3            95 ( 5           0.00094 ( 0.00023
  H245F           89 ( 6           830 ( 30          0.098 ( 0.002              Km values for luciferin and ATP,2 possibly as a consequence
  H245A           15 ( 2           240 ( 30          0.52 ( 0.06                of having Phe at position 244 (P. pyralis numbering).
  H245Q           19 ( 3           380 ( 28          0.027 ( 0.003                 Lysine at a position equivalent to 529 of WT is absolutely
  H245N           15 ( 2           354 ( 23          0.3 ( 0.02
  H245D           64 ( 5          1820 ( 40          0.0037( 0.0003             conserved in all luciferases2,3 (12-18) and members of the
  K529R           16 ( 2            40 ( 10          0.0026 ( 0.0001            adenylating superfamily (24). The importance of a cationic
  K529Q          161 ( 13          829 ( 129         0.0015 ( 0.0001            group at position 529 for productive substrate binding is
  T343S           27 ( 4           385 ( 69          0.358 ( 0.05
  T343A           99 ( 17          857 ( 361         0.0145 ( 0.001             confirmed by the deleterious effects on the Km values for
    a                                                                           both luciferin and ATP observed with the K529Q, but not
      Kinetic constants were determined as described in ref 25. b The
kcat values were calculated by dividing the Vmax values, obtained from          the K529R, mutants (Table 2). Moreover, the kcat values of
the data acquired for the determination of the corresponding Km values          the Lys529 mutants were greatly reduced (Table 2), indicat-
for luciferin, by the concentration of the luciferases present in the assays.   ing a likely functional role for this residue. Additional studies
                                                                                on the role of Lys529 in luciferase are in progress.
                                                                                   In WT, changes at residues His245 and Thr343 affect
                                                                                bioluminescence color. His245 is absolutely conserved in
                                                                                all luciferases, and our model predicts (25) that the side-
                                                                                chain imidazole is at the opening of the luciferin binding
                                                                                pocket in close proximity to both the luciferin carboxylate
                                                                                and the γ-phosphate of ATP. The imidazole group appears
                                                                                to be in a position to remove the C-4 proton of luciferin
                                                                                (Figure 1), a critical first step in the enzyme-catalyzed
                                                                                oxidation of the substrate. The kcat value of the H245A
                                                                                enzyme lacking the side chain, however, is reduced only
                                                                                3-fold compared to WT. This result does not support a major
                                                                                role for the His245 residue in luciferase catalysis. Further,
                                                                                the kinetic properties of the H245A protein (Table 2) suggest
                                                                                that His245 is not necessary for productive luciferin binding
                                                                                and that this residue plays only a minor role in Mg-ATP
                                                                                binding and catalysis of bioluminescence. The steady-state
                                                                                kinetic data (Table 2) for the new His245 mutants reported
FIGURE 3: Bioluminescence emission spectra of WT and the Thr343                 here (H245R, H245Q, and H245N) support these and earlier
mutants at pH 7.8. Conditions for recording and correcting the                  conclusions (25). Attractive and repulsive ion pair interac-
emission spectra are described under Materials and Methods.                     tions with the γ-phosphate of ATP likely account for the
                                                                                decreased and increased Km values for Mg-ATP of the
   Bioluminescence Emission. The bioluminescence emission                       H245R and H245D proteins, respectively. Possibly, the
spectra of the luciferases were measured at pH 5.5 and 7.8,                     considerably lower specific activities of the Arg and Asp
and the measured emission maxima are tabulated in Table                         mutants also result from ion pair interactions interfering with
1. The corrected and normalized bioluminescence spectra                         the release of pyrophosphate during the formation of the
obtained at pH 7.8 for WT and for the His245 and Thr343                         luciferyl-adenylate (Figure 1).
mutants are shown in Figures 2 and 3. The colors of light                          Our model of the P. pyralis luciferase active site (25)
emitted from alkaline and acidic reactions of WT, as well                       predicted that both the main-chain NH and side-chain OH
as both of the K529 mutants, H244F and T343S, were almost                       groups of Thr343 form hydrogen bonds to the R-phosphate
identical (Table 1 and Figure 3), being green at the higher                     of ATP. Identical interactions between the equivalent residue
pH and red at low pH. In contrast, T343A emitted only red                       (Thr326) and AMP are observed in the crystal structure of
light at pH 7.8. The His245 mutants produced broadened                          PheA (24). The T343A mutant lacking the side-chain
emission spectra with increasing maximal emission wave-                         hydroxyl group binds both enzyme substrates much more
lengths at pH 7.8 from 579 to 613 nm with replacements                          poorly than WT and H245A (Table 2). Also, the kcat value
Arg < Phe < Ala < Gln < Asn (Table 1 and Figure 2).                             of T343A is ∼110-fold lower than WT and ∼35-fold lower
The Asp replacement produced only red light. All of the                         than H245A. Apparently, Thr343 has a more significant role
mutants and WT emit red light at pH 5.5.                                        than His245 in both luciferase substrate binding and catalysis.
DISCUSSION                                                                      The similar kinetic constants of H245A and the conservative
                                                                                T343S mutant (Table 2) further substantiate the greater
   Roles of His244, His245, Thr343, and Lys529 in P. pyralis                    relative importance of Thr343. We speculate that in the
Luciferase Bioluminescence. Luciferases with mutations at                       superfamily of adenylating enzymes, threonine at the position
positions 244 and 529 catalyzed light emission nearly
identical in color to WT at both alkaline and acidic pH (Table                    2 Monika Gruber, Promega Corp., Madison, WI, personal com-
1). It is therefore very unlikely that either His244 or Lys529                  munication.
is involved in bioluminescence color determination. The                           3 V. R. Viviani, personal communication.
13228 Biochemistry, Vol. 38, No. 40, 1999                                                                          Branchini et al.
equivalent to 343 in P. pyralis luciferase participates in
orientating ATP for reaction with the carboxylate substrate,
and perhaps, in conjunction with Lys529, stabilizes the key
pentavalent transition state of the adenylation step.
   Firefly Bioluminescence Color: General. Bioluminescence
catalyzed by P. pyralis luciferase in vivo and at pH 7.8 in
vitro is yellow-green with an emission maximum of ∼560
nm. Differences in bioluminescence color are caused by the
following: natural species variations in luciferase structure3
(7, 8, 12, 18); amino acid substitutions introduced by
mutagenesis techniques (12-18); in vitro substitutions of
analogues of luciferin and ATP for the natural substrates (3,
5, 30, 35-37); alkylation of cysteine (and possibly other)
residues (38, 39); and modifications of the in vitro assay
conditions, including the addition of certain divalent metal
ions (e.g., Hg2+ and Zn2+), increasing temperature above ∼25
°C, and maintaining pH in the range ∼5 to ∼7 (3). The
specific amino acid(s) responsible for the pH sensitivity of
the firefly luciferases is (are) not known.
   Luciferases2,3 (7, 8, 12, 18) from three families of beetles
use identical substrates to generate light ranging in maximum
emission wavelength from ∼530 to ∼635 nm. Alterations
in luciferase tertiary structure may affect bioluminescence
color by causing changes in the polarity of the active site,      FIGURE 4: Bioluminescence emission spectra of WT over the pH
misalignment of an essential base, thereby disrupting its         range 5.5-9.0. Conditions for recording and correcting the emission
tautomerization function (Figure 1), or disruption of interac-    spectra are described under Materials and Methods.
tions essential to the maintenance of the proper C2-C2′
dihedral angle of the emitter (Figure 1). It is likely that two   ion (Figure 2). Thus, the hydrogen bonding ability of the
distinct excited state emitters produce green light (∼540 nm)     Nδ1 and N2 positions alone is not sufficient for green light
and red light (∼630 nm) in beetle bioluminescence. The            emission. The His245 side chain may be necessary to fix
various colors observed in nature might then be caused by         the positions of at least two different groups, and this point
structural variations in the proteins that affect the precise     is discussed further below.
ground state-excited state energy differences of the two             Although the location of His245 in our model (25) suggests
emitters. Bioluminescence emission spectra of the fireflies,      it might be the luciferase base required in White’s (5, 6)
but not the click beetles and railroad worms, are likely          keto-enol mechanism (Figure 1), this is quite unlikely since
comprised of contributions from both of the emitters, as is       the emission spectra of the H245R and H245F mutants, both
apparently the case for WT (Figure 4 and Table 1).                lacking nucleophilic functional groups, contain significant
   Roles of His245 and Thr343 in Bioluminescence Color.           amounts of green light (Table 1 and Figure 2). Moreover, if
Both His245 and Thr343 are absolutely conserved in the 20         the His245 side-chain group were essential, one would expect
known luciferase sequences2,3 (12-18) including 12 from           the bioluminescence of the H245A mutant, lacking any of
the firefly, 5 click beetles, and 3 “glow worms”. Among these     the steric or functional properties of the His side chain, to
sequences, one click beetle isozyme catalyzes orange light        be completely red-shifted; however; this was not observed.
emission (12, 18), and the railroad worm enzyme produces             His245 is at an N-cap position of a helix that comprises
red light3. The bioluminescence emission spectra obtained         part of the luciferin binding site in our model (25). Green
from the series of His245 mutants reported here as arranged       light emission appears to be favored by a positive charge at
in Figure 2 resemble the broadened red-shifted WT spectra,        position 245 and/or the maintenance of the helix dipole, as
as the assay pH is adjusted from 7.8 to 5.5 (Figure 4). The       well as by side-chain groups providing substantial nonpolar
protonation of His245 at acidic pH, however, cannot be the        area (Table 1 and Figure 2). An electropositive region of
direct cause of the red-shifted WT emissions, since all           the emitter site would be expected to interact favorably with
luciferases from the three beetle families have His at position   the enolate form of oxyluciferin, a feature supporting the
245 and they are not affected by pH. Moreover, the least          theory advanced by White (5, 6). However, a properly
altered emission spectrum (λmax ) 579 nm) obtained with           oriented electropositive center could also fix the keto form
the His245 mutants was that of H245R which contains the           of the emitter in a higher energy TICT excited state
positively charged guanidinium group. Possibly, the imida-        conformation favoring green light emission according to
zolium ion form of His245 may actually favor green light          McCapra’s proposal (10, 11). The substantially green emis-
emission.                                                         sion from the H245F mutant, containing the large aromatic
   The H245Q and H245N mutants with side-chain amides             side chain, may be a consequence of steric restrictions
isosteric for the N2 and Nδ1 of His, respectively, were          similarly fixing the conformation of the emitter. The observed
prepared to examine the potential functions of these sites.       color differences among the His245 mutants cannot be
The emission spectrum produced by H245Q was similar to            accounted for by any single property (40) of the amino acids
that of H245A which lacks the imidazole side chain, while         replacing His245. The side-chain imidazole group at position
the H245N-catalyzed emission closely resembled that of            245, however, appears to be required for normal green light
H245D which contains the negatively charged carboxylate           emission in the firefly.
Firefly Bioluminescence Color                                                           Biochemistry, Vol. 38, No. 40, 1999 13229

FIGURE 5: Schematic representation of the speculative mechanisms for bioluminescence color determination. The figure on the left indicates
the participation of His245, Thr343, AMP, and Arg218 in maintaining a specific conformation of the keto form of excited state oxyluciferin
through hydrogen bonding interactions (||||). The tautomerization of the keto form of oxyluciferin to the corresponding enol emitter is
indicated by the arrows.

FIGURE 6: Stereo diagram showing the interactions suggested by molecular modeling of P. pyralis luciferase with oxyluciferin and AMP.
Traces through the R-carbons of regions N197-G200, H244-T252, V217-F219, S314-L319, and Q338-I351 are shown as coils. This
diagram was generated using the program MOLSCRIPT (42).

   The T343S and T343A mutants catalyzed the emission of                McCapra’s proposal (10, 11). The participation of His245
yellow-green and red light (Figure 3), respectively, and the            and Arg218, residues such as Thr343 that are absolutely
corresponding emission spectra were nearly identical to those           conserved among all luciferases, and AMP is indicated in
of WT at pH values 7.8 and 5.5 (Figure 4). Apparently, a                the speculative mechanisms (Figure 5). We propose that
side-chain hydroxyl group at position 343 is absolutely                 Arg218 anchors the benzothiazole ring of the emitter through
essential for normal yellow-green light emission in the                 an ion pair interaction with the phenolate ion (Figures 5 and
firefly. In Figure 5, we present a speculative schematic                6), an interaction identified previously for luciferin and this
diagram that illustrates how Thr343 may function as an                  luciferase residue (25). His245 may be required to properly
essential determinant of yellow-green light emission. The               orient both the emitter and Thr343 through hydrogen bonding
approximate relative positions of AMP, the keto form of                 interactions between the His side-chain group and both the
oxyluciferin, and the side chains of His245, Thr343, and                carbonyl of oxyluciferin and a phosphate oxygen of AMP.
R218 shown in Figure 5 are consistent with molecular                    The hydrogen bonding proton donors could both be from
modeling results (Figure 6) that extend our previous studies            the imidazolium form (Figure 5), or, alternatively, from the
(25). The side-chain hydroxyl of Thr343 may serve as the                neutral form and a bridging water molecule. AMP is included
required functional group for the abstraction of the luciferin          as a key component fixing the position of the Thr343 side
C-5 proton according to White’s mechanism (5, 6) as                     chain, enabling it to make an essential hydrogen bond to
indicated by the arrows in Figure 5. Alternatively, the role            the emitter carbonyl and possibly removing the C-5 proton.
of Thr343 (Figure 5) may be to provide an essential hydrogen            The location of AMP in close proximity to Thr343, His245,
bond to the keto form of the emitter to maintain a TICT                 and the carbonyl of the emitter is suggested from modeling
excited state conformer that emits green light as required in           studies (Figure 6). Based on studies with ATP analogues
13230 Biochemistry, Vol. 38, No. 40, 1999                                                                             Branchini et al.
(35-37), the inclusion of AMP is warranted because of the                nescence: Fundamentals and Applied Aspects (Campbell, A.
likelihood that it remains bound to luciferase both while the            K., Kricka, L. J., and Stanley, P. E., Eds.) pp 387-391, John
emitter is generated and during the radiative decay process.             Wiley and Sons, Chichester.
                                                                     11. McCapra, F. (1997) in Bioluminescence and Chemilumines-
   Concluding Remarks. The results of mutational and                     cence: Molecular Reporting with Photons (Hastings, J. W.,
biochemical studies reported here have enabled us to propose             Kricka, L. J., and Stanley, P. E., Eds.) pp 7-15, John Wiley
speculative mechanisms for color determination in firefly                and Sons, Chichester.
bioluminescence. An essential role for Thr343, the participa-        12. Wood, K. V. (1995) Photochem. Photobiol. 62, 662-673.
tion of His245 and Arg218, and the involvement of bound              13. Sala-Newby, G. B., and Campbell, A. K. (1991) Biochem. J.
AMP are indicated. The proposal can account for the                      279, 727-732.
                                                                     14. Kajiyama, N., and Nakano, E. (1991) Protein Eng. 4, 691-
enzyme’s role in color determination according to either of              693.
the major theories of White and McCapra. Unfortunately,              15. Ohmiya, Y., Hirano, T., and Ohashi, M. (1996) FEBS Lett.
our data cannot be used to distinguish these theories.                   384, 83-86.
Although threonine hydroxyl groups (pK ∼17) do not                   16. Mamaev, S. V., Laikhter, A. L., Arslan, T., and Hecht, S. M.
commonly function as enzyme bases, our mechanism estab-                  (1996) J. Am. Chem. Soc. 118, 7243-7244.
lishes favorable geometry for a tautomerism process requiring        17. Arslan, T., Mamaev, S. V., Mamaeva, N. V., and Hecht, S.
                                                                         M. (1997) J. Am. Chem. Soc. 119, 10877-10887.
little displacement of the atoms involved. Moreover, the             18. Wood, K. V., Lam, Y. A., Seliger, H. H., and McElroy, W.
tautomerism of oxyluciferin should be quite facile based on              D. (1989) Science 244, 700-702.
the reported (41) extensive enolization of 5-methyloxylu-            19. Ye, L., Buck, L. M., Schaeffer, H. J., and Leach, F. R. (1997)
ciferin in neutral methanol. We believe that White’s mech-               Biochim. Biophys. Acta 1339, 39-52.
anism (5, 6) remains viable; however, the simple requirement         20. Suzuki, H., Kawarabayasi, Y., Kondo, J., Abe, T., Nishikawa,
of tight hydrogen-bonding interactions to maintain normal                K., Kimura, S., Hashimoto, T., and Yamamoto, T. (1990) J.
                                                                         Biol. Chem. 265, 8681-8685.
emission conditions according to the McCapra theory (10,             21. Babbitt, P. C., Kenyon, G. L., Martin, B. M., Charest, H.,
11) is attractive. Furthermore, we believe that the keto form            Sylvestre, M., Scholten, J. D., Chang, K.-H., Liang, P.-H., and
of oxyluciferin, the red light emitter in both mechanisms,               Dunaway-Mariano, D. (1992) Biochemistry 31, 5594-5604.
exists in the TICT excited state. In nature, the color variations    22. Conti, E., Franks, N. P., and Brick, P. (1996) Structure 4, 287-
found among beetles may be modulated mainly by changes                   298.
in the anisotropic polarity of the emitter site, and residues        23. Chang, K.-H., Xiang, H., and Dunaway-Mariano, D. (1997)
                                                                         Biochemistry 36, 15650-15659.
contributing significantly to this property likely include           24. Conti, E., Stachelhaus, T., Marahiel, M. A., and Brick, P.
His245 and Arg218. For the extreme red-emitting luciferases              (1997) EMBO J. 16, 4174-4183.
of the click beetle (12, 18) and railroad worm,3 structural          25. Branchini, B. R., Magyar, R. A., Murtiashaw, M. H., Anderson,
variations in the oxyluciferin binding site may position the             S. M., and Zimmer, M. (1998) Biochemistry 37, 15311-15319.
emitter so that the functions of Thr343 and/or His245 in             26. Franks, N. P., Jenkins, A., Conti, E., Lieb, W. R., and Brick,
producing yellow-green light are no longer possible. Ad-                 P. (1998) Biophys. J. 75, 2205-2211.
                                                                     27. Branchini, B. R., Magyar, R. A., Marcantonio, K. M.,
ditional mutational studies are currently in progress to better          Newberry, K. J., Stroh, J. G., Hinz, L. K., and Murtiashaw,
understand the role of specific firefly luciferase residues in           M. H. (1997) J. Biol. Chem. 272, 19359-19364.
the bioluminescence process.                                         28. Branchini, B. R., Magyar, R. A., Murtiashaw, M. H., Mag-
                                                                         nasco, N., Hinz, L. K., and Stroh, J. G. (1997) Arch. Biochem.
ACKNOWLEDGMENT                                                           Biophys. 340, 52-58.
                                                                     29. Thompson, J. F., Geoghegan, K. F., Lloyd, D. B., Lanzetti,
   We thank Monika Gruber and Justin Stroh for helpful                   A. J., Magyar, R. A., Anderson, S. M., and Branchini, B. R.
discussions; David Lloyd, John Thompson, and the DNA                     (1997) J. Biol. Chem. 272, 18766-18771.
Sequencing Facility, Pfizer Inc., for providing the DNA              30. Branchini, B. R., Hayward M. M., Bamford, S., Brennan P.
sequencing data; Boris Chrunyk for assistance with the                   M., and Lajiness, E. J. (1989) Photochem. Photobiol. 49, 689-
circular dichroism studies; and Luke Ward, Jud Grice, and                695.
Evelyn Bamford for technical assistance.                             31. Herbst, R., Schäfer, U., and Seckler, R. (1997) J. Biol. Chem.
                                                                         272, 7099-7105.
REFERENCES                                                           32. Zoller, M. J., and Smith, M. (1983) Methods Enzymol. 100,
                                                                         468-500.
  1. Lloyd, J. E. (1978) in Bioluminescence in Action (Herring, P.   33. Green, A. A., and McElroy, W. D. (1956) Biochim. Biophys.
     J., Ed.) pp 241-272, Academic Press, New York.                      Acta 20, 170-176.
  2. Hastings, J. W. (1995) in Cell Physiology Source Book           34. Laemmli, U. K. (1970) Nature 227, 680-685.
     (Sperelakis, N., Ed.) pp 665-681, Academic Press, New York.     35. Lee, R., Denburg, J. L., and McElroy, W. D. (1970) Arch.
  3. DeLuca, M. (1976) AdV. Enzymol. 44, 37-68.                          Biochem. Biophys. 141, 38-52.
  4. Seliger, H. H., and McElroy, W. D. (1960) Arch. Biochem.        36. DeLuca, M., Leonard, N. J., Gates, B. J., and McElroy, W.
     Biophys. 88, 136-141.                                               D. (1973) Proc. Natl. Acad. Sci. U.S.A. 70, 1664-1666.
  5. White, E. H., Rapaport, E., Seliger, H. H., and McElroy, W.     37. Rosendahl, M. S., Leonard, N. J., and DeLuca, M. (1982)
     D. (1971) Bioorg. Chem. 1, 92-122.                                  Photochem. Photobiol. 35, 857-861.
  6. White, E. H., Steinmetz, M. G., Miano, J. D., Wildes, P. D.,    38. White, E. H., and Branchini, B. R. (1975) J. Am. Chem. Soc.
     and Morland, R. (1980) J. Am. Chem. Soc. 102, 3199-3208.            97, 1243-1245.
  7. McElroy, W. D., and Seliger, H. H. (1966) in Molecular          39. Alter, S. C., and DeLuca, M. (1986) Biochemistry 25, 1599-
     Architecture in Cell Physiology (Hayashi, H., and Szent-            1605.
     Gyorgyi, I., Eds.) pp 63-79, Prentice-Hall, Englewood Cliffs,   40. Creighton, T. E. (1993) Proteins, 2nd ed., Chapter 4, W. H.
     NJ.                                                                 Freeman and Company, New York.
  8. Hastings, J. W. (1996) Gene 173, 5-11.                          41. White, E. H., and Roswell, D. F. (1991) Photochem. Photobiol.
  9. Morton, R. A., Hopkins, T. A., and Seliger, H. H. (1969)            53, 131-136.
     Biochemistry 8, 1598-1607.                                      42. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946-950.
 10. McCapra, F., Gilfoyle, D. J., Young, D. W., Church, N. J.,
     and Spencer, P. (1994) in Bioluminescence and Chemilumi-           BI991181O
You can also read